weka.classifiers.timeseries.core

Interface StateDependentPredictor

All Superinterfaces:

weka.classifiers.Classifier

public interface StateDependentPredictor
extends weka.classifiers.Classifier

An interface for state-dependent predictors. Has methods for clearing, setting and getting previous model state as well as for model/state serialization. Useful for forecasters which use models which store state of last prediction to be used in the next prediction. A state-dependent predictor is Created by pedrofale on 17-08-2016.

Method Summary

Methods

Modifier and Type	Method and Description
void	clearPreviousState() Clear/reset state of the model.
Object	getPreviousState() Get the last set state of the model.
void	loadSerializedState(String path) Load serialized model state
void	serializeState(String path) Serialize model state
void	setPreviousState(Object previousState) Load state into model.

Methods inherited from interface weka.classifiers.Classifier

buildClassifier, classifyInstance, distributionForInstance, getCapabilities

Method Detail

clearPreviousState

void clearPreviousState()

Clear/reset state of the model.

setPreviousState

void setPreviousState(Object previousState)

Load state into model.

Parameters:

previousState - the state to set the model to

getPreviousState

Object getPreviousState()

Get the last set state of the model.

Returns:

the state of the model to be used in next prediction

serializeState

void serializeState(String path)
                    throws Exception

Serialize model state

Parameters:

path - the path to the file to hold the serialized state

Throws:

Exception

loadSerializedState

void loadSerializedState(String path)
                         throws Exception

Load serialized model state

Parameters:

path - the path to the file to load the serialized state from

Throws:

Exception